Berthelot's coherence and overconvergence conjecture for relative rigid cohomology

Let (S,S,S)(S,S,\mathcal{S}) be a triple endowed with a lift of Frobenius on S\mathcal{S}, and let

(X,X)f(Y,Y)g(S,S)(X,\overline{X}) \xrightarrow{f} (Y,\overline{Y}) \xrightarrow{g} (S,S)

be a diagram of pairs such that f:XYf:\overline{X}\to\overline{Y} is proper, f1(Y)=Xf^{-1}(Y)=X, and fX:XYf|_X:X\to Y is smooth. Let E\mathcal{E} be an overconvergent (F-)(F\text{-})isocrystal on (X,X)/SK(X,\overline{X})/\mathcal{S}_K, and let q0q\geq 0. For the subcategory C\mathcal{C} of (Y,Y)(Y,\overline{Y})-triples over (S,S,S)(S,S,\mathcal{S}) specified by the conjecture, write I((Z,Z)/SK,Z)I^{\dagger}((Z,\overline{Z})/\mathcal{S}_K,\mathcal{Z}) for the category of overconvergent isocrystals over Z\mathcal{Z}, and let pip_i denote the projections from the realization over Z×SZ\mathcal{Z}\times_{\mathcal{S}}\mathcal{Z} to that over Z\mathcal{Z}. Berthelot's conjecture. There exists a subcategory C\mathcal{C} such that there exists uniquely an overconvergent (F-)(F\text{-})isocrystal F\mathcal{F} on (Y,Y)/SK(Y,\overline{Y})/\mathcal{S}_K whose restriction to every (Z,Z,Z)C(Z,\overline{Z},\mathcal{Z})\in\mathcal{C} with Z\mathcal{Z} formally smooth over S\mathcal{S} near ZZ is functorially given by

(Rqf(X×YZ,X×YZ)/Z,rigE,ϵ),\left(R^qf_{(X\times_Y Z,\overline{X}\times_{\overline{Y}}\overline{Z})/\mathcal{Z},\operatorname{rig}*}\mathcal{E},\epsilon\right),

where ϵ\epsilon is the canonical pair of base-change isomorphisms through Z×SZ\mathcal{Z}\times_{\mathcal{S}}\mathcal{Z}. This weaker version is intended to be sufficient for the unique existence of the qq-th rigid-cohomology overconvergent isocrystal, while the stronger version requires the condition for every suitable triple.

Sources & referencesView supporting material

Primary source

Atsushi Shiho, “Relative log convergent cohomology and relative rigid cohomology II”, arXiv:0707.1743 (2008).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.